A variety of physical and chemical carcinogens have been demonstrated to induce ras gene mutations in both human and animal tumor models. It has recently been reported that ras mutation-positive adult acute myeloid leukemia (AML) is strongly associated with specific occupational exposures. The overall objective of this Childrens Cancer Group (CCG) project is to define the relationship between reported and measured environmental exposures and the presence or absence of ras gene mutations in leukemic blasts from a large population of children (n=244O) diagnosed with acute lymphoblastic leukemia (ALL) and AML. Utilizing the extensive exposure data collected through two CCG case-control epidemiologic investigations and by conducting molecular studies of proto-oncogenes K- ras and N-ras in malignant blast cells from diagnostic bone marrow specimens, it will be possible to test the following hypotheses: (1) In utero and/or post-natal exposure to specific environmental agents are associated with an increased risk of ras p21 mutation in childhood ALL and AML. Environmental agents to be tested include: solvents, petroleum products, pesticides, medications, alcohol, tobacco products, x-rays, low frequency electromagnetic field radiation and radon; (2) Associations identified differ according to the morphologic/biologic features of the leukemia; (3) The mutational spectra of ras mutations is specific for individual environmental exposures; and (4) Cases positive for ras gene mutation have a poorer event free-survival than cases negative for ras gene mutation, independent of other known prognostic factors. K-ras and N-ras gene mutations in codons 12, 13, and 61 from bone marrow cells will be measured using material available from the CCG Biopathology Center, Leukemia Reference Laboratories and primary treatment institutions. Statistical analyses will be performed to compare the environmental exposures of cases with ras mutation to those cases without ras mutation (case-case comparison) and to compare both case groups to their initially matched normal controls (case-matched control comparison) as well as to the control group as whole (case-pooled control comparison). The proposed investigation of ras gene mutations in childhood leukemia, utilizing the unique resources available through the CCG, will provide important data regarding ras p21 mutations as a potential biologic indicator of exposure to specific environmental agents in childhood acute leukemia cases. Analysis of the mutational spectra of ras proto-oncogenes, in conjunction with the environmental exposure data, may provide new insights into the molecular etiology of childhood acute leukemias.